The Mathematics Methods Research In Denoising And The Whole Domain Denoising Method Using Singular Value Decomposition

Image denoising is important for the subsequent image processing, and mathematics plays an irreplaceable part in image processing. Based on the comprehension of related references, this paper reviews the important mathematical denoising methods. It introduces the basic theory of these methods and their improvement, and also points out their advantages and weakness in denoising.Especially, the SVD methods used to denoise are presented in the paper. Then a new method (SVDWD) is proposed in allusion to the disadvantage of the traditional methods which only deal with some of the singular values. The new method presents the singular value difference function and estimates the formal singular values according to the function, then restores the noisy image with the estimated singular values.In the end, the paper puts forward the prospect of mathematical methods in image denoising.